Seed-Induced Uniqueness in Transformer Models: Subspace Alignment Governs Subliminal Transfer

Ayşe Selin Okatan, Mustafa İlhan Akbaş, Laxima Niure Kandel +1

We analyze subliminal transfer in Transformer models, where a teacher embeds hidden traits that can be linearly decoded by a student without degrading main-task performance. Prior…